Download Polynucleotide Phosphorylase and Mitochondrial

yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Lipid signaling wikipedia , lookup

Biochemistry wikipedia , lookup

Light-dependent reactions wikipedia , lookup

Biosynthesis wikipedia , lookup

Microbial metabolism wikipedia , lookup

Metalloprotein wikipedia , lookup

Amino acid synthesis wikipedia , lookup

Metabolism wikipedia , lookup

Enzyme inhibitor wikipedia , lookup

Electron transport chain wikipedia , lookup

Arsenic biochemistry wikipedia , lookup

Enzyme wikipedia , lookup

Glycolysis wikipedia , lookup

NADH:ubiquinone oxidoreductase (H+-translocating) wikipedia , lookup

Evolution of metal ions in biological systems wikipedia , lookup

Citric acid cycle wikipedia , lookup

Adenosine triphosphate wikipedia , lookup

Glutathione wikipedia , lookup

Mitochondrial replacement therapy wikipedia , lookup

Oxidative phosphorylation wikipedia , lookup

Mitochondrion wikipedia , lookup

TOXICOLOGICAL SCIENCES 117(2), 270–281 (2010)
Advance Access publication May 10, 2010
Polynucleotide Phosphorylase and Mitochondrial ATP Synthase Mediate
Reduction of Arsenate to the More Toxic Arsenite by Forming
Arsenylated Analogues of ADP and ATP
Balázs Németi,* Maria Elena Regonesi,† Paolo Tortora,† and Zoltán Gregus*,1
*Department of Pharmacology and Pharmacotherapy, Toxicology Section, University of Pécs, Medical School, H-7624 Pécs, Hungary; and †Department of
Biotechnologies and Biosciences, University of Milano-Bicocca, I-20126 Milan, Italy
To whom correspondence should be addressed at Department of Pharmacology and Pharmacotherapy, Toxicology Section, University of Pécs, Medical School,
Szigeti út 12, H-7624 Pécs, Hungary. Fax: þ36-72-536-218; E-mail: [email protected]
Received March 30, 2010; accepted May 1, 2010
We have demonstrated that phosphorolytic-arsenolytic enzymes
can promote reduction of arsenate (AsV) into the more toxic
arsenite (AsIII) because they convert AsV into an arsenylated
product in which the arsenic is more reducible by glutathione
(GSH) or other thiols to AsIII than in inorganic AsV. We have also
shown that mitochondria can rapidly reduce AsV in a process
requiring intact oxidative phosphorylation and intramitochondrial
GSH. Thus, these organelles might reduce AsV because mitochondrial ATP synthase, using AsV instead of phosphate,
arsenylates ADP to ADP-AsV, which in turn is readily reduced
by GSH. To test this hypothesis, we first examined whether the
RNA-cleaving enzyme polynucleotide phosphorylase (PNPase),
which can split poly-adenylate (poly-A) by arsenolysis into units
of AMP-AsV (a homologue of ADP-AsV), could also promote
reduction of AsV to AsIII in presence of thiols. Indeed, bacterial
PNPase markedly facilitated formation of AsIII when incubated
with poly-A, AsV, and GSH. PNPase-mediated AsV reduction
depended on arsenolysis of poly-A and presence of a thiol. PNPase
can also form AMP-AsV from ADP and AsV (termed arsenolysis
of ADP). In presence of GSH, this reaction also facilitated AsV
reduction in proportion to AMP-AsV production. Although
various thiols did not influence the arsenolytic yield of AMPAsV, they differentially promoted the PNPase-mediated reduction
of AsV, with GSH being the most effective. Circumstantial
evidence indicated that AMP-AsV formed by PNPase is more
reducible to AsIII by GSH than inorganic AsV. Then, we
demonstrated that AsV reduction by isolated mitochondria was
markedly inhibited by an ADP analogue that enters mitochondria
but is not phosphorylated or arsenylated. Furthermore, inhibitors
of the export of ATP or ADP-AsV from the mitochondria
diminished the increment in AsV reduction caused by adding
GSH externally to these organelles whose intramitochondrial
GSH had been depleted. Thus, whereas PNPase promotes
reduction of AsV by incorporating it into AMP-AsV, the
mitochondrial ATP synthase facilitates AsV reduction by forming
ADP-AsV; then GSH can easily reduce these arsenylated
nucleotides to AsIII.
Key Words: arsenate; reduction; ATP synthase; polynucleotide
phosphorylase; glutathione; arsenolysis.
Arsenic is one of the oldest poisons known to humans.
Widely distributed in Earth’s crust, inorganic arsenicals dissolve
from the bedrock into water, thereby ultimately causing
contamination of food and drinking water. Prolonged arsenic
exposure is associated with a wide variety of human diseases,
including skin, vascular, and neurological disorders and cancer
(Brinkel et al., 2009; Hughes et al., 2007; Platanias, 2009;
Tseng, 2007). The pentavalent arsenate (AsV), the prevalent
form of inorganic arsenic in the environment, enters the body via
the intestinal inorganic phosphate (Pi) transporters (VillaBellosta and Sorribas, 2008). Inside the cells, AsV can impair
energy homeostasis by replacing Pi in enzyme reactions, owing
to the structural similarity of these oxyanions (Hughes, 2006).
AsV, however, can also be reduced to the trivalent arsenite
(AsIII) in a glutathione (GSH)-dependent fashion (Csanaky and
Gregus, 2005). The importance of this metabolic conversion is
underlined by the following two facts: (1) AsIII is much more
toxic than AsV because it exhibits strong covalent reactivity
toward SH-groups, especially vicinal dithiols, thus inactivating
essential proteins (Dilda and Hogg, 2007; Kitchin and Wallace,
2008); (2) reduction of AsV to AsIII is the indispensable first
step preceding the production of pentavalent and trivalent
methylated metabolites, among which the latter are even more
toxic than AsIII (Stýblo et al., 2002; Thomas, 2007).
The toxicological significance of AsV reduction has fuelled
intensive research to explore the proteins and mechanisms
involved. Many microorganisms have evolved specific AsVreducing enzymes, which, when coupled to AsIII exporters,
confer resistance to arsenic toxicity (Messens and Silver, 2006;
Páez-Espino et al., 2009; Rosen, 2002). In mammals, however,
enzymes specific for reducing AsV have not been found.
Nevertheless, several enzymes have been demonstrated to
The Author 2010. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved.
For permissions, please email: [email protected]
facilitate the conversion of AsV to AsIII in presence of their
respective substrate and a thiol compound. These include purine
nucleoside phosphorylase (Gregus and Németi, 2002; Németi
and Gregus, 2009b; Radabaugh et al., 2002), glyceraldehyde-3phosphate dehydrogenase (Gregus and Németi, 2005), glycogen phosphorylase (Gregus and Németi, 2007; Németi and
Gregus, 2007), ornithine carbamoyl transferase (OCT) (Németi
and Gregus, 2009a), and the bacterial enzyme phosphotransacetylase (Németi and Gregus, 2009b). Each of these enzymes
utilizes Pi for cleaving the substrate (termed phosphorolysis)
into two compounds, one of which is a phosphorylated
metabolite. When AsV substitutes for Pi, arsenolysis is carried
out and an arsenylated metabolite is formed. We have recently
provided evidence that these phosphorolytic-arsenolytic
enzymes do not directly reduce AsV to AsIII, but rather they
convert AsV into an arsenylated metabolite (i.e., an AsV ester
or anhydride), in which the pentavalent arsenic is much more
reducible by thiol compounds than in the inorganic form
(Gregus et al., 2009; Németi and Gregus, 2009b).
Isolated rat liver mitochondria can also rapidly reduce AsV
(Németi and Gregus, 2002). Although the contributing enzymes have not been identified, mitochondrial AsV reduction was
shown to be sensitive to agents that inhibit oxidative phosphorylation and chemicals that deplete mitochondrial GSH
(Németi and Gregus, 2002). Interestingly, ATP synthase that
phosphorylates ADP to ATP can use not only Pi but also AsV.
In the latter case, however, ADP-AsV, a purportedly unstable
analogue of ATP, is formed (Chan et al., 1969; Gresser, 1981;
Moore et al., 1983). Thus, it might be hypothesized that
mitochondrial ATP synthase facilitates conversion of AsV to
AsIII by arsenylating ADP to ADP-AsV, which may then be
readily reduced by GSH, just like the arsenylated products of
the phosphorolytic-arsenolytic enzymes listed previously.
The aim of this study was to examine the previously stated
hypothesis. However, ATP synthase can carry out the synthesis
of ATP or ADP-AsV only in functionally and structurally
intact mitochondria, precluding direct testing of this hypothesis
on the isolated enzyme. Therefore, to assess the assumption
that ADP-AsV formation in mitochondria may underlie the
mechanism of AsV reduction by these organelles, we took a
twofold experimental approach. First, we evaluated our hypothesis using a model enzyme, polynucleotide phosphorylase
(PNPase). PNPase is an evolutionarily conserved 3# / 5#
exoribonuclease that can phosphorolytically degrade RNA into
nucleoside diphosphate units and the homopolymeric RNA
variant poly-adenylate (poly-A) into ADP (Littauer and Soreq,
1982; Sarkar and Fisher, 2006). When AsV substitutes for Pi,
PNPase catalyzes the arsenolytic decay of poly-A, forming
AMP-AsV (Fig. 1). Considering the PNPase-produced AMPAsV as a homolog of the ATP synthase–generated ADP-AsV,
we first examined whether or not PNPase that forms AMP-AsV
promoted reduction of AsV by thiols, and if it did, whether
AsV reduction was based on the formation of AMP-AsV.
Thereafter, using mitochondria isolated from rat liver, we
sought for circumstantial evidence that ADP-AsV formation is
of importance in mitochondrial AsV reduction. To this end, we
studied the responsiveness of mitochondrial AsV reduction to
various chemical probes that selectively affect the formation or
fate of ATP or ADP-AsV. These included (1) an ADP analogue, which is taken up but is not phosphorylated or arsenylated in mitochondria, (2) known inhibitors of the adenine
nucleotide translocator (ANT) that can block the export of ATP
or ADP-AsV from the matrix space, and (3) inhibitors of ATP
synthase. In order to facilitate our experimental inquiry, some
studies were performed on mitochondria with their GSH
content depleted.
1-chloro-2,4-dinitrobenzene (CDNB), 2-mercaptoethanol (2-ME), dithiothreitol (DTT), dimercaptopropane-1-sulfonic acid (DMPS), meso-dimercaptosuccinic
acid (DMSA), ADP, 2#-deoxyadenosine-5#-diphosphate (dADP), poly-adenylate
(poly-A), bongkrekic acid (BKA), carboxyatractyloside (CATR), aurovertin B
(AU), a-b-methylene-ADP (me-ADP), rabbit muscle pyruvate kinase (PK),
lactate dehydrogenase (LDH), isopropyl-b-D-thiogalactopyranoside (IPTG),
PMSF, ampicillin, DNase I, and RNase were from Sigma. GSH, NADH,
oligomycin (OM), and disodium hydrogen arsenate (AsV) were from Reanal Ltd
(Budapest, Hungary). Protease inhibitor tablets were from Roche Diagnostics
(Mannheim, Germany); Ni-NTA His bind resin was from Novagen; phosphoenolpyruvic acid lithium salt (PEP) was from Fluka; and sodium arsenite (AsIII)
was purchased from Carlo Erba (Milan, Italy). Cyclosporine A was a generous
gift from Novartis (Basel, Switzerland). The sources of chemicals used in arsenic
speciation and for the enzymatic analysis of GSH have been given elsewhere
(Csanaky et al., 2003; Németi and Gregus, 2002). All other chemicals were of the
highest purity commercially available.
Experiments with PNPase
PNPase purification and assay. His6-tagged wild-type PNPase was
purified from Escherichia coli GF5322 strain as described previously
(Matus-Ortega et al., 2007). Cultures were grown up to an optical density of
0.8 at 600 nm in 100 ml of LD-broth with 100 lg/ml ampicillin at 30C with
shaking. Expression of PNPase was induced by adding 0.5mM IPTG and, after
further incubation for 3 h at the same temperature, cells were collected by
centrifugation, washed with 50mM Tris-HCl, pH 7.4. Cells were resuspended
in 5 ml/g cell of lysis buffer (50mM Tris-HCl, pH 8.0; 0.5 M NaCl; 5mM
imidazole, pH 8.0; 1mM PMSF; 5% glycerol; and 1 tablet/50 ml solution of
protease inhibitors) and disrupted by sonication at 0C. The lysate was
incubated with DNase I (5 lg/ml) and RNase A (10 lg/ml) for 30 min at 0C.
The sample was clarified by centrifugation at 20,000 3 g for 30 min at 4C.
Potassium phosphate, pH 8.0, was then added to the supernatant to a final
concentration of 10mM. The resulting solution was dialyzed for 60 min at 37C
against 500 ml of lysis buffer containing 10mM potassium phosphate, pH 8.0.
Finally, the dialyzed sample was centrifuged at 15,000 3 g for 10 min at 4C
and the supernatant applied to a Ni-NTA His bind column (bed volume, 1 ml),
pre-equilibrated with 10 volumes of lysis buffer. The column was washed with
10 volumes of the same buffer, and PNPase was eluted with a discontinuous
gradient of imidazole (1 volume of 0.05, 0.1, 0.2, 0.4 M imidazole). Fractions
were stored at 20C in 50% glycerol.
PNPase from E. coli was assayed by the spectrophotometric method of
Ghetta et al. (2004), which records the decrease in NADH concentration (at
340 nm) during the PNPase-limited formation of ADP from poly-A in presence
of PEP, PK, and LDH. In this assay, PK converts ADP to ATP while using PEP
and producing pyruvate, which is ultimately reduced by LDH to lactate at the
FIG. 1. PNPase and ATP synthase form homologous arsenylated nucleotides, AMP-AsV and ADP-AsV, respectively. These arsenylated compounds are
purportedly unstable because they undergo hydrolysis to produce AsV plus AMP or ADP, respectively. It should be noted that PNPase can also catalyze the
arsenolysis of ADP, which also yields AMP-AsV.
expense of NADH. The assay mixture contained (in a final volume of 1 ml)
100mM Tris-HCl buffer (pH 7.7 at room temperature), 0.2 mg/ml poly-A,
5mM MgCl2, 1mM PEP, 0.2mM NADH, 1.3 U/ml PK, 10 U/ml LDH, and the
appropriately diluted PNPase. The reaction was started by adding 10mM
potassium phosphate. One unit is the amount of PNPase that forms 1 lmole
ADP in 1 min.
Assaying PNPase for thiol-dependent AsV-reducing activity. AsVreducing activity of purified PNPase was assayed in 100mM Tris buffer (pH
7.7 at room temperature) containing 5mM MgCl2. The enzyme was preincubated with poly-A (typically 0.2 mg/ml) and a thiol compound (typically
GSH at 10mM) at 37C for 5 min, then AsV (200lM) was added to start the
30-min incubation in a final volume of 0.3 ml. Variations from these conditions
are specified in Figures 2–5.
To study the PNPase-mediated thiol-dependent AsV reduction mechanistically, four types of incubations (type A–D) were performed. For clarity, the
design of these incubations is presented in a table within Figure 6. These
experiments were designed to test the hypotheses that AsV reduction takes
place not only when the arsenolysis of poly-A (that forms AMP-AsV) proceeds
in presence of GSH (type A incubations) but also when GSH is added only after
AMP-AsV had been formed by the arsenolysis, which was terminated by dADP
plus Pi just before GSH addition (type D incubations); however, AsV reduction
is abolished when formation of AMP-AsV is completely prevented by dADP
plus Pi even if GSH is added later (type C). Other incubations were performed
for comparative purpose, i.e., to test the effect on AsV reduction of the delayed
addition of GSH as in type D, but without terminating the arsenolysis before
GSH addition (type B), and to correct for nonenzymatic AsV reduction by
performing incubations lacking PNPase. The timing of these experiments and
the concentrations of PNPase, substrates (i.e., poly-A and AsV), and inhibitors
of the arsenolytic reaction (i.e., dADP and Pi) are given in Figure 6.
All incubations were stopped by sequential addition of 100 ll of 50mM
CdSO4 solution followed by 100 ll of 1.5 M perchloric acid solution containing 50mM HgCl2. The rationale for this procedure has been given elsewhere (Németi and Gregus, 2004). The incubates thus treated were stored
at 80C until arsenic analysis.
Because AsV is reduced nonenzymatically in presence of a thiol to a small
extent, the nonenzymatic AsIII formation rates were regularly determined from
incubations lacking PNPase but containing the appropriate thiol compound and
subtracted from the rates measured when the incubation contained both enzyme
and thiol. The enzymatic AsV-reducing activity thus calculated was expressed
as the amount of AsIII formed per minute and unit PNPase.
Assaying PNPase for arsenolytic and AsV-reducing activities. In order to
compare the AsV-reducing activity (i.e., AsIII formation from AsV) and
arsenolytic activity (i.e., AMP formation from poly-A) of the enzyme, purified
E. coli PNPase (50 mU/ml) was preincubated at 37C with poly-A (0.2 mg/ml)
and one of the thiol compounds (GSH, 2-ME, DMSA, DMPS, or DTT; at
10mM SH-group concentration) in 100mM Tris buffer (pH 7.7) containing
5mM MgCl2 for 5 min, as described previously. Then, AsV (200lM) was
added to start the 30-min incubation. Immediately before the incubations were
stopped, the incubates were divided into two samples. In the sample to be used
for arsenic analysis, AsV reduction was stopped as described previously. In the
sample destined for AMP analysis, the arsenolysis was terminated by adding
1/6 volume of 2.1 M perchloric acid solution. Other details are given in
Figure 5. A similar general experimental procedure was used when PNPase was
tested for arsenolysis of ADP and the coupled AsV reduction in presence of
GSH. For details, see Figure 4.
Experiments with Rat Liver Mitochondria
Isolation of mitochondria. Male Wistar rats kept under standardized
conditions and weighing 270–290 g were obtained from the SPF breeding
house of the University of Pécs (Hungary). All procedures were carried out on
animals according to the Hungarian Animals Act, and the study was approved
by the Ethics Committee on Animal Research of the University of Pécs.
Rat liver mitochondria were isolated as described previously (Németi and
Gregus, 2002). All steps were carried out at 0–4C. Briefly, the rat was
euthanized and its liver was homogenized manually with a Potter-Elvehjem
glass-Teflon homogenizer in ice-cold isolation buffer (250mM sucrose, 5mM
Tris, and 1mM ethylene glycol-bis(2-amino-ethylether)-N,N,N’,N’-tetraacetic
acid (EGTA), pH 7.2 at room temperature). The homogenate (~120 ml) was
then centrifuged at 500 3 g for 5 min. The obtained supernatant was
centrifuged at 5000 3 g for 15 min. The resultant pellet containing the
mitochondria from a single liver was resuspended in the isolation buffer
(~100 ml) and centrifuged again (5000 3 g, 15 min). This washing procedure
FIG. 2. PNPase-mediated reduction of AsV—dependence on the concentrations of PNPase, AsV, poly-A, and GSH. Escherichia coli PNPase (50 mU/ml
or as indicated in panel A) was preincubated at 37C for 5 min with GSH
(10mM or as given in panel D) and poly-A (0.2 mg/ml or as shown in panel C)
in 100mM Tris buffer (pH 7.7) containing 5mM MgCl2. Then, AsV (200lM or
as shown in panel B) was added to start the incubation lasting for 30 min.
Symbols represent AsIII formation rates (mean ± SEM) in three incubations.
was repeated once more. The final mitochondrial pellet was resuspended in ~1
ml of isolation buffer. This procedure yields mitochondria virtually free from
contaminating soluble proteins and cell organelles (Gregus et al., 1998) and of
excellent functionality (Németi and Gregus, 2002). The protein concentration
of the mitochondrial suspension was determined by means of the biuret method
(Gornall et al., 1949) and usually was 90–100 mg/ml.
Depletion of mitochondrial GSH. Procedures published by others (Han
et al., 2003; Jocelyn and Cronshaw, 1985) were adapted to deplete mitochondria of GSH with CDNB. Briefly, isolated mitochondria suspended in the
isolation buffer at 10 mg/ml protein concentration were incubated with CDNB
(100lM; dissolved in ethanol) for 2 min at room temperature in presence of
cyclosporine A (10lM) to prevent permeability transition. At the end of the
incubation, mitochondria were centrifuged at 8000g, 2C for 5 min. After
discarding the supernatant, the pelleted mitochondria were resuspended in icecold isolation buffer and centrifuged again to remove the excess CDNB. The
resultant pellet was gently resuspended, and its protein concentration was
determined. Control mitochondria were processed similarly but were incubated
with ethanol (0.25%) instead of CDNB. Mitochondria thus prepared were used
for assaying AsV-reducing activity within 1 h.
The GSH concentration in the perchloric acid-deproteinized supernatants of
untreated, ethanol-treated, and CDNB-treated mitochondria was determined by
the method of Tietze (1969). The GSH concentration of untreated mitochondria
(7.85 ± 0.53 nmol/mg protein; n ¼ 5) decreased insignificantly by ethanol
treatment to 6.99 ± 0.18 nmol/mg and markedly by CDNB treatment to 0.22 ±
0.02 nmol/mg (n ¼ 5 per group). At ~3 h after preparation of mitochondria,
mitochondrial O2 consumption supported by the respiratory substrates
FIG. 3. PNPase-mediated reduction of AsV—effects of compounds
countering the PNPase-catalyzed arsenolysis of poly-A. Escherichia coli
PNPase (50 mU/ml) was preincubated at 37C for 5 min with GSH (10mM)
and poly-A (0.2 mg/ml) in 100mM Tris buffer (pH 7.7) containing 5mM
MgCl2. Then, ADP or dADP or phosphate at the indicated concentrations plus
AsV (200lM) were added to start the 30-min incubation. Symbols represent
AsIII formation rates (mean ± SEM) in three incubations. With the exception of
the AsIII formation rate in presence of ADP at 1mM, all values are significantly
different (p < 0.05) from the AsIII formation rate observed in absence of
glutamate (5mM) and malate (1mM) was measured with a Clark-electrode
before and after addition of ADP as previously described (Németi and Gregus,
2002). The mitochondrial respiration and respiratory control ratio remained
unchanged in response to ethanol or CDNB treatment.
Assaying mitochondrial AsV-reducing activity. AsV-reducing activity of
isolated rat liver mitochondria was assayed in a 10mM hydroxyethyl piperazine
ethane sulfonic acid (HEPES) buffer containing 120mM KCl and 1mM EGTA
(pH 7.4) at 37C. As the incubation conditions (i.e., the concentrations of
mitochondrial protein and reagents, as well as the time and termination mode of
incubations) varied in these experiments, specific details of these assays are
described in Figures 8–10. AsIII formation rates were corrected for proteinindependent AsV reduction by performing incubations lacking mitochondria.
Analytical Procedures
Arsenic analysis. After having been subjected to protein precipitation as
described previously, the incubates from the AsV reductase assays were
centrifuged at 10,000 3 g, 4C for 10 min. AsIII and AsV in the resultant
supernatants were separated and quantified by high-performance liquid
chromatography–hydride generation–atomic fluorescence spectrometry, using
a strong anion exchange guard column and an analytical column (both
Hamilton PRP X-100) and eluted isocratically with 60mM sodium phosphate
buffer (pH 5.75). The details of this analysis have been given elsewhere
(Gregus et al., 2000; Németi et al., 2003).
AMP analysis. AMP is the final product of the arsenolysis of poly-A or
ADP by PNPase because the immediate product AMP-AsV undergoes
spontaneous hydrolysis. In the supernatants of the deproteinized incubates,
AMP was quantified by a high-performance liquid chromatography analysis.
Briefly, the samples were injected through a Rheodyne 7125 injector equipped
with a 20-ll sample loop onto a 3.9 X 20-mm guard column followed by a
FIG. 4. Effects of ADP concentration on the PNPase-mediated arsenolysis
of ADP (as measured by AMP production) and AsV reduction (as measured by
AsIII formation). Escherichia coli PNPase (50 mU/ml) was preincubated at
37C for 5 min with GSH (10mM) in 100mM Tris buffer (pH 7.7) containing
5mM MgCl2. Then, ADP (at concentrations indicated) and AsV (200lM) were
added in rapid succession to start the 30-min incubation. Bars represent AMP
or AsIII formation rates (mean ± SEM) in three incubations.
3.9 3 50-mm analytical column (both Novapack C18, particle size 4 lm), and
eluted isocratically with 100mM sodium phosphate buffer (pH 5.4) at a flow
rate of 1 ml/min. The absorbance of the effluent was monitored at 254 nm by
a flow-through spectrophotometer (Waters 486) and recorded by Millenium 3.2
(Waters, Milford, MA). AMP eluted at 6.1 min. Quantification was based on
peak areas of samples and authentic standards.
Statistics. Data were analyzed using one-way ANOVA followed by
Duncan’s test or Student’s t-test, with p < 0.05 as the level of significance.
FIG. 5. Effects of thiols on the PNPase-mediated arsenolysis of poly-A (as
measured by AMP production) and AsV reduction (as measured by AsIII
formation). Escherichia coli PNPase (50 mU/ml) was preincubated at 37C for
5 min with poly-A (0.2 mg/ml) and the thiol compound indicated (at 10mM
SH-group) in 100mM Tris buffer (pH 7.7) containing 5mM MgCl2. Then, AsV
(200lM) was added to start the 30-min incubation. Bars represent AMP or
AsIII formation rates (mean ± SEM) in three incubations.
Testing PNPase for Mediating Thiol-Supported AsV
In order to determine if PNPase could facilitate the thioldependent reduction of AsV coupled to arsenolysis of poly-A,
we incubated PNPase with AsV in presence of its arsenolytic
substrate poly-A and GSH. These incubations revealed that
PNPase promoted reduction of AsV in presence of GSH and
poly-A but not in absence of either GSH or poly-A. AsV
reduction increased linearly with the incubation time up to
60 min (not shown) and the enzyme concentration (Fig. 2A),
and asymptotically with the concentrations of AsV (Fig. 2B)
and poly-A (Fig. 2C), reaching the maximal AsIII formation
rate at 0.2 mg/ml poly-A. GSH (1–15mM) increased PNPasemediated AsV reduction in a concentration-dependent sigmoid
fashion (Fig. 2D).
To further characterize the PNPase-facilitated AsV reduction,
we investigated the effects of compounds, which can interfere
with the arsenolysis of poly-A by this enzyme. These included
dADP (an inhibitor of PNPase), Pi (the substrate of PNPase
competing with AsV), and ADP (the product of PNPase-
catalyzed phosphorolysis of poly-A). As demonstrated in Figure 3,
dADP and Pi exerted a concentration-dependent inhibition, which
was significant (p < 0.05) even at a concentration as low as
0.25mM, and both dADP and Pi abolished AsV reduction at
4mM. Interestingly, however, ADP enhanced AsIII formation
significantly at concentrations below 1mM and inhibited AsV
reduction only at concentrations of 2mM and above.
Next, we examined the possibility that the increased AsV
reduction at low ADP concentrations (Fig. 3) is because of
PNPase-catalyzed arsenolysis of ADP (i.e., formation of AMPAsV from ADP). It has been shown that the PNPase can also
promote an exchange reaction between the terminal phosphate
of nucleoside diphosphates and Pi (Godefroy et al., 1970;
Littauer and Soreq, 1982). Thus, ADP arsenolysis would take
place when AsV is used instead of Pi. To test this hypothesis,
we incubated PNPase with ADP (0.25–1mM) and AsV in
presence of GSH but without poly-A and quantified the amount
of AMP produced (as an index of AMP-AsV formation) and
the amount of AsIII formed (as an index of AsV reduction). As
demonstrated in Figure 4, no AMP or AsIII was produced in
absence of ADP, whereas in presence of 0.25mM ADP,
FIG. 6. Mechanistic studies on PNPase-mediated thiol-dependent AsV reduction—formation of AsIII is most significant when the PNPase-catalyzed arsenolysis
of poly-A takes place in presence of a thiol (incubations labeled with A and B); however, it also occurs when the arsenolysis is permitted transiently then terminated
chemically and the thiol is added thereafter (incubation D) but is negligible when the arsenolysis is chemically prevented (incubation C). The general incubation
schedule is depicted in the table above. Each incubation contained Escherichia coli PNPase (250 mU/ml), poly-A (0.2 mg/ml), and 100mM Tris buffer (pH 7.7)
containing 5mM MgCl2 and was started, after preincubation at 37C for 5 min, by adding AsV (200lM) at time 0. Into incubation A (A0-5 and A0-8) GSH (10mM),
whereas into incubation C, the inhibitors (INH) of the arsenolytic reaction, i.e., dADP (5mM) plus Pi (10mM), were added immediately before AsV. Into incubations
B, C, and D, GSH was added at 5 min, whereas into incubation D, the INH was added immediately before GSH. At times indicated in the table above, the
incubations were stopped with thiol reagents (see the ‘‘Materials and Methods’’ section), and AsIII formed in the incubates was quantified. Bars represent the
amounts of AsIII formed (mean ± SEM) in three incubations. The value of A5-8 represents the difference in average AsIII formation in incubations A0-8 and A0-5.
considerable amount of AsIII and AMP were formed.
Interestingly, however, increasing the concentrations of ADP
to 0.5 and 1mM resulted in gradual declines in the rates of both
AsIII and AMP formation, as compared with those seen at
0.25mM ADP. Thus, the AMP and AsIII formation rates
changed proportionately to each other, with the former
exceeding the latter more than 10-fold (Fig. 4).
In order to further analyze the relationship between the
arsenolysis of poly-A and the reduction of AsV to AsIII, we
compared the rate of the two coupled reactions either in
absence or in presence of various thiols (i.e., GSH, DMSA,
DTT, DMPS, and 2-ME). Figure 5 (top) shows that the rate of
PNPase-catalyzed arsenolysis of poly-A into AMP-AsV was
completely independent of the presence or absence of any thiol
compound. In contrast, AsV reduction could not be observed in
absence of thiols (Fig. 5, bottom). The various thiol compounds
differentially enhanced AsV reduction, with GSH exhibiting
the strongest support. It is noteworthy that even in the presence
of GSH the rate of AMP formation (Fig. 5, top) exceeded
several fold the rate of AsIII production (Fig. 5, bottom).
Because findings demonstrated in Figures 4 and 5 may
indicate that PNPase forms AMP-AsV, which in turn is reduced
readily by a thiol to AsIII, we tested the assumption that arsenic
in AMP-AsV produced in the PNPase-catalyzed arsenolysis of
poly-A is more reducible to AsIII by GSH than in inorganic
AsV. For this purpose, we carried out four types of experiments,
as summarized in the table in Figure 6. In type C incubations,
PNPase was inhibited by dADP plus Pi from the start, thus
AMP-AsV formation was prevented. In these assays, a very
small amount of AsIII could be detected despite the presence of
GSH in the last 3 min (Fig. 6). In type D incubations, arsenolysis of poly-A was permitted in absence of GSH, therefore
AMP-AsV could be formed but it could not react with the thiol.
Then, AMP-AsV production was stopped by adding dADP plus
Pi, immediately followed by addition of GSH, which then could
reduce arsenic in the preformed AMP-AsV. In type D incubations, four times as much AsIII (Fig. 6) was found than in
type C experiments (p < 0.05). Type A and type B incubations
were performed for comparative purpose. In type A incubations,
the arsenolysis of poly-A took place in the presence of GSH for
5 or 8 min, and bars A0-5 and A0-8, respectively, indicate the
amount of AsIII produced in these incubations. It is interesting
to note that AsIII formation was much higher in the last 3 min of
A0-8 incubations (bar labeled A5-8) than the amount of AsIII
found in the 8-min long type B incubations, in which AMPAsV was formed in absence of GSH for the first 5 min, but in
presence of GSH in the last 3 min.
Testing Mitochondrial ATP Synthase for Mediating
Thiol-Supported AsV Reduction
In experiments on isolated rat liver mitochondria, we sought
for circumstantial evidence that ATP synthase promotes thioldependent AsV reduction by forming ADP-AsV, a hypothesis
depicted in Figure 7. First, we examined whether AsV
reduction by mitochondria supplied with GSH externally was
inhibited by the ANT blocker CATR or BKA. These experiments were performed on both intact mitochondria, containing
endogenous GSH, and on mitochondria whose GSH content
had been depleted. As Figure 8 demonstrates, GSH-containing
intact mitochondria reduced AsV at a relatively high rate,
which could neither be increased significantly with exogenous
GSH nor be diminished by the ANT inhibitor BKA or CATR,
irrespective of the presence of exogenous GSH (left panel). In
contrast, the markedly lower AsV-reducing capacity of GSHdepleted mitochondria (~15% of GSH-containing mitochondria) was increased threefold by addition of GSH (Fig. 8,
FIG. 7. Hypothetical mechanism of AsV reduction supported by mitochondrial ATP synthase. ADP and AsV transported into mitochondria by the ANT and
the phosphate transporter (PiT), respectively, are coupled to form ADP-AsV by ATP synthase (ASY). ADP-AsV may then be reduced by intramitochondrial GSH,
or after export from the mitochondrion through ANT, by extramitochondrial GSH to produce AsIII and ADP. Alternatively, ADP-AsV may undergo hydrolysis
into ADP and AsV. Note that AsV uptake by mitochondria may also be mediated by the dicarboxylate carrier (not shown). CATR and BKA are inhibitors of ANT,
whereas OM and AU are inhibitors of ASY.
right panel). Importantly, both ANT inhibitors significantly
diminished the amount of AsV reduced by GSH-depleted
mitochondria supplied with GSH externally.
Secondly, we performed experiments with intact mitochondria using methylene-ADP (me-ADP), which enters mitochondria via the ANT in exchange for intramitochondrial adenine
nucleotides but is not phosphorylated by ATP synthase (Duée
and Vignais, 1969). Figure 9 demonstrates that preincubation
of mitochondria with me-ADP resulted in a concentrationdependent diminution of mitochondrial formation of AsIII from
AsV that was significant (p < 0.05) even at the lowest me-ADP
concentration (25lM) and was near complete at 250lM
FIG. 8. AsV reduction by intact liver mitochondria—effects of the ANT inhibitor BKA and CATR on the increment in AsIII formation caused by addition of GSH
to control and GSH-depleted mitochondria. To deplete the endogenous GSH from mitochondria, isolated rat liver mitochondria were treated with CDNB dissolved in
ethanol, whereas control mitochondria were subjected to ethanol treatment, as described in detail in the ‘‘Materials and Methods’’ section. The thus-obtained GSHdepleted and GSH-containing mitochondria (2 mg protein/ml) were preincubated at 37C for 5 min in KCl-HEPES buffer containing glutamate (5mM) and cyclosporine
A (2lM). Thereafter, an ANT inhibitor (10lM BKA or 5lM CATR) or buffer, GSH (10mM) or buffer, and AsV (50lM) were added in rapid succession to start the
5-min incubation in a final volume of 0.3 ml. The incubations were stopped by sequential addition of 100 ll of 50mM CdSO4 solution containing 1% Triton X-100
followed by 100 ll of 1.5 M perchloric acid solution containing 50mM HgCl2. Bars represent AsIII formation (mean ± SEM) in three incubations with mitochondria
prepared from different rats. Asterisks indicate AsIII formation rates significantly lower (p < 0.05) than observed in the absence of ANT inhibitor.
FIG. 9. AsV reduction by mitochondria—effects of me-ADP, an ADP
analogue that mitochondria cannot phosphorylate or arsenylate. Isolated rat
liver mitochondria (2 mg protein/ml) were preincubated at 37C for 5 min with
me-ADP (at the indicated concentrations) in KCl-HEPES buffer containing
glutamate (5mM) and cyclosporine A (2lM). Thereafter, ADP (0, 250, or
500lM) and AsV (50lM) were added in rapid succession to start the 5-min
incubation in a final volume of 0.3 ml. The incubations were stopped by
sequential addition of 100 ll of 50mM CdSO4 solution containing 1% Triton
X-100 followed by 100 ll of 1.5 M perchloric acid solution containing 50mM
HgCl2. Symbols represent AsIII formation (mean ± SEM) in three incubations
with mitochondria prepared from different rats. Asterisks indicate AsIII
formation rates significantly different (p < 0.05) from those observed in the
absence of ADP.
me-ADP. This inhibition of AsV reduction could be alleviated
considerably by 0.25 or 0.5mM ADP added 5 min after meADP. Interestingly, ADP added at these concentrations
produced a weak, though statistically significant, decrease in
AsIII formation in absence of me-ADP.
Thirdly, we carried out studies on GSH-depleted mitochondria that had been preloaded with ADP and glutamate. These
incubations were scheduled as shown in the table inserted in
Figure 10. Type A incubations were started with AsV plus one
of the inhibitors of ANT (CATR or BKA), in order to retain the
ADP-AsV produced inside the mitochondria. Three minutes
after AsV addition, the ATP synthase inhibitor AU and the
detergent Nonidet P-40 (NiP) were added to prevent the ADPAsV from being hydrolyzed by ATP synthase (which works in
the reverse mode when the mitochondrial inner membrane is
disrupted) and to solubilize mitochondria, respectively. NiP
was added into the incubations either alone or together with
GSH. The incubations were stopped 3 min later. In type A
incubations, ~2 nmol/ml AsIII was formed in presence of either
CATR or BKA and in absence of exogenous GSH (Fig. 10).
Addition of exogenous GSH to such incubations increased the
reduction of AsV by 40%. In type B incubations, the
mitochondrial ATP synthase was inhibited from the start by
AU (type B1 incubations) or OM (type B2 incubations), thus
FIG. 10. Mechanistic studies on AsV reduction by GSH-depleted
mitochondria—formation of AsIII is significant when ADP-AsV production
is permitted transiently then terminated with the ATP synthase inhibitor AU
plus a detergent (NiP) and GSH is added thereafter (incubation A with GSH)
but is much less when ADP-AsV production is prevented by the ATP synthase
inhibitor AU or OM (incubations B1 and B2). The general incubation schedule
is depicted in the table above. In all experiments, CDNB-pretreated GSHdepleted mitochondria (10 mg protein/ml) were preincubated with the
respiratory substrate glutamate (5mM) and ADP (4mM) in presence of
cyclosporine A (2lM) at 37C for 3 min. Thereafter, incubation A was started
at time 0 by adding the adenine nucleotide translocator inhibitor (ANTi) BKA
(50lM; lower panel) or CATR (25lM; upper panel) immediately followed by
AsV (100lM), whereas incubations B1 and B2 were started at time zero by
adding an ANTi, an ATP synthase inhibitor (10lM AU in B1 incubations and
100lM OM in B2 incubations) and AsV in rapid succession. Three minutes
after AsV, GSH (10mM) and the detergent Nonidet P-40 (NiP; 0.1%) were
added into incubations B1 and B2, whereas AU followed by GSH and NiP were
added into incubation A. For comparison, all types of incubations were also
carried out without addition of GSH at 3 min. The final incubation volume was
0.3 ml. All incubations were stopped at 6 min with sequential addition of 100 ll
of 50mM CdSO4 solution followed by 100 ll of 1.5 M perchloric acid solution
containing 50mM HgCl2. Bars represent the amounts of AsIII formed (mean ±
SEM) in three incubations with mitochondria prepared from different rats.
ADP-AsV formation was prevented. Irrespective of which ATP
synthase inhibitor was used in these experiments, AsIII
formation rates were approximately one third of that seen in
the respective type A incubations, in which ADP-AsV could be
formed (Fig. 10). GSH increased AsIII formation rates in these
incubations as well; however, the GSH-induced increment in
AsIII production was larger (1.2 nmol/ml) with uninhibited
ATP synthase (incubation A) than with inhibited ATP synthase
(0.4 nmol/ml; incubations B1 and B2).
This work has been devised to evaluate the hypothesis that
mitochondria can reduce AsV to AsIII via ATP synthase, the
enzyme that converts ADP to ATP in the process of oxidative
phosphorylation, utilizing the energy of the inwardly directed
Hþ gradient generated by the electron transport chain through
oxidation of reduced coenzymes. This postulation was based
on three pieces of information. First, we reported some time
ago that rat liver mitochondria, working like chemical reactors,
take up AsV, rapidly reduce it to AsIII, and then export the
formed AsIII (Németi and Gregus, 2002). Importantly,
mitochondrial reduction of AsV was found to be accelerated
by glutamate, whose oxidation yields reduced coenzymes and
thus fuels oxidative phosphorylation; however, it was decreased or virtually halted by agents that impair oxidative
phosphorylation, such as inhibitors of the electron transport
complexes, chemicals dissipating the Hþ gradient, and/or the
mitochondrial membrane potential and thus arresting the Hþ
influx that propels ATP synthase, inhibitors of the FO proton
channel and ATP synthase, ATP, and detergents that disrupt
membrane integrity. In addition to its apparent dependence on
oxidative phosphorylation, mitochondrial AsV reduction also
depends on the availability of GSH in these organelles because
it was markedly impaired in hepatic mitochondria isolated from
rats pretreated with GSH depletors (Németi and Gregus, 2002).
Secondly, it has long been known that mitochondrial ATP
synthase, when supplied with AsV instead of Pi, can produce
ADP-AsV instead of ATP (Chan et al., 1969; Gresser, 1981;
Moore et al., 1983). Third, we have recently demonstrated that
several phosphorolytic-arsenolytic enzymes (which can cleave
their substrates with Pi or AsV) can facilitate reduction of AsV
in presence of a thiol, such as GSH, because in the arsenolytic
process they form an arsenylated metabolite, which is exquisitely susceptible to reduction by a thiol into AsIII (Gregus
et al., 2009). Therefore, it appeared plausible that ATP synthase can promote AsV reduction by GSH in a similar fashion,
i.e., through arsenylating ADP.
The hypothesis that phosphorolytic-arsenolytic enzymes
promote reduction of AsV has been verified directly, i.e., by
using the purified enzymes. This approach, however, could not
be adopted for testing whether or not ATP synthase mediates
AsV reduction by arsenylating ADP because this enzyme when
isolated loses its driving force, the transmembrane proton
influx, and becomes unable to phosphorylate or arsenylate
ADP. Nevertheless, this paper presents two lines of circumstantial evidence that ATP synthase can promote reduction of
AsV to AsIII by producing ADP-AsV in which the pentavalent
arsenic is more reducible by GSH than in the inorganic AsV.
For evaluation of the above-stated hypothesis, we could
fortuitously use purified PNPase as a surrogate of ATP
synthase because this enzyme can produce AMP-AsV,
a homologue of ADP-AsV, by arsenolytic cleavage of poly-A,
a synthetic RNA analogue (Littauer and Soreq, 1982; Fig. 1). It
was found that PNPase isolated from E. coli can indeed
mediate reduction of AsV to AsIII in presence of GSH in direct
proportion to the enzyme concentration (Fig. 2). Moreover, its
AsV-reducing activity depended on both the presence of GSH
or other thiols (Figs. 2 and 5) and the arsenolysis of poly-A,
i.e., the formation of AMP-AsV. This latter conclusion is
supported by observations indicating that the AsV-reducing
activity of PNPase (1) is negligible in absence of poly-A and
increases with elevation of poly-A concentration (Fig. 2) and
(2) is inhibited by compounds that are known to counteract the
arsenolysis of poly-A, such as dADP (Bon et al., 1970; Chou
et al., 1975), Pi, and ADP at high concentrations (Fig. 3).
Curiously, ADP at low concentrations promoted PNPasemediated AsV reduction (Fig. 3). When explaining the
mechanism of this concentration-dependent effect of ADP, it
should be noted that ADP enters into two mutually exclusive
reactions catalyzed by PNPase, thereby exerting opposing
influences on AMP-AsV formation, and in turn, on AsV reduction. As the PNPase-catalyzed reaction is readily reversible
(Littauer and Soreq, 1982), ADP can undergo polymerization
into oligo-adenylate and/or can support elongation of poly-A,
thereby countering arsenolysis of poly-A by the enzyme and
diminishing AMP-AsV production. However, PNPase can also
catalyze an exchange reaction, which results in arsenolysis of
ADP in presence of AsV, yielding AMP-AsV plus Pi
(Godefroy et al., 1970; Littauer and Soreq, 1982). Indeed,
PNPase at low ADP concentration in absence of poly-A
yielded significant amounts of AMP-AsV (as indicated by
AMP formation) and mediated AsV reduction in presence of
GSH (Fig. 4). At higher ADP concentrations, however, less
and less AMP-AsV and AsIII was formed, suggesting that
ADP increasingly entered into the polymerization reaction
rather than arsenolytic AMP-AsV formation. Importantly, the
proportionality of AMP-AsV production (as quantified by
formation of AMP, the hydrolytic product of AMP-AsV) and
AsV reduction clearly observed in these studies (Fig. 4) further
supports the conclusion that PNPase-mediated GSH-dependent
AsV reduction occurs via formation of AMP-AsV.
Although our findings discussed above indicate that the
arsenolytic activity (i.e., AMP-AsV formation) is a prerequisite
for the facile PNPase-mediated AsV reduction, the arsenolytic
AMP-AsV production is clearly not the sole determinant of the
rate of the PNPase-mediated thiol-supported AsV reduction.
This conclusion is evident from observations shown in Figure 5,
which demonstrates that the rate of PNP-catalyzed arsenolysis
(as quantified by AMP formation) is similar irrespective of the
presence (or absence) of any of the thiol compounds tested,
whereas the rate of AsV reduction measured by AsIII formation
varies markedly in presence of various thiol compounds.
A similar observation was made with respect to purine
nucleoside phosphorylase, another phosphorolytic-arsenolytic
enzyme, then prompting the assumption that it forms its
arsenylated product (i.e., ribose-1-AsV), which is effectively but
differentially reduced by various thiols (Németi and Gregus,
2009b), a hypothesis subsequently supported by experimental
and theoretical evidence (Gregus et al., 2009). The present work
provides circumstantial evidence that the arsenylated product of
PNPase-catalyzed arsenolysis of poly-A (i.e., AMP-AsV) is
more effectively reduced to AsIII by GSH than inorganic AsV.
This conclusion can be deduced from Figure 6 that demonstrates
that more AsV was reduced when the experimental conditions
permitted the preformation of AMP-AsV from poly-A and AsV
by PNPase before adding GSH (incubation D), as compared
with the condition when AMP-AsV could not be produced by
PNPase from poly-A and AsV before adding GSH because the
arsenolytic activity of PNPase had been inhibited by dADP and
Pi (incubation C). Yet, most of the formed AMP-AsV produced
by PNPase in absence of GSH (i.e., in 0–5 min of incubation D)
must have been hydrolyzed and thus unavailable for reduction
by GSH upon addition of this thiol. This is indicated by the
huge amount of AsIII formed when PNPase could produce
AMP-AsV in presence of GSH (i.e., in incubations A0-5). Even
in presence of GSH (or another thiol), most AMP-AsV produced by PNPase must undergo hydrolysis, rather than thiolmediated reduction, as revealed by the much higher rate of AMP
formation (Figs. 4 and 5, top) than AsIII production (Figs. 4 and
5, bottom).
The second line of circumstantial evidence for ATP synthase
mediating reduction of AsV to AsIII via producing ADP-AsV
comes from experiments on isolated mitochondria. These
studies were designed to seek for support of the model of
mitochondrial AsV reduction presented in Figure 6. The results
of these studies are compatible with the proposed model at
several points. For example, the reduction of AsV in mitochondria was nearly abolished by preincubation of these
organelles with me-ADP (Fig. 9), which can be taken up by
mitochondria through the ANT in exchange for intramitochondrial adenine nucleotides (Duée and Vignais, 1969) but cannot
be phosphorylated by ATP synthase, although it does not
inhibit phosphorylation of ADP (Jones and Boyer, 1969).
Thus, the impaired AsV reduction in response to me-ADP is
likely explained by replacement of the intramitochondrial ADP
pool with this ADP analogue, which cannot be arsenylated to
form me-ADP-AsV, thereby preventing formation of a product
that is efficiently reducible by GSH to AsIII. This interpretation
is supported by the finding that the me-ADP-induced impairment of AsV reduction was largely reversed by adding ADP
(Fig. 9). Even the observation that ADP addition to the mitochondria slightly but significantly decreased the AsV reduction
in absence of me-ADP (Fig. 9) is compatible with the hypothetic model in Figure 7. While inside the matrix space, ADP
should promote ADP-AsV production by ATP synthase;
however, ADP added externally may also have a converse
influence. This may result from import of the external ADP
through the ANT into the matrix, which must occur with
simultaneous export of internal ADP-AsV because ANT works
as an exchanger of ADP and ATP (Klingenberg, 2008) and
purportedly of ADP and ADP-AsV. Therefore, the increased
efflux of ADP-AsV from the matrix space abundant in GSH
into the GSH-free incubation medium could have diminished
the amount of ADP-AsV available for reaction with GSH and
reduction to AsIII. It has been shown that ADP lowers arsenic
concentration in corn mitochondria preloaded with AsV and
concluded that this effect involves efflux of AsV as ADP-AsV
(Bertagnolli and Hanson, 1973).
The experiments with isolated mitochondria subjected to
DNCB-induced depletion of GSH indicating a 90% decrease in
AsV reduction as compared with the GSH-containing mitochondria (Fig. 8) confirm the contention that mitochondria
require GSH for reducing AsV to AsIII (Németi and Gregus,
2002). The studies on the GSH-depleted mitochondria to which
GSH was added externally further signify the importance of
both GSH and formation of ADP-AsV in AsV reduction by
mitochondria because GSH addition caused an increment in
formation of AsIII (Fig. 8, right); however, this increment was
significantly diminished by either BKA or CATR, highly
selective and potent ANT inhibitors targeting this translocator
from opposite membrane sides (Klingenberg, 2008). In congruence with the model (Fig. 7), these findings may suggest
that normally ADP-AsV is largely reduced to AsIII inside the
matrix space by endogenous GSH; however, a fraction of
ADP-AsV is exported by ANT and may be reduced outside the
matrix, provided GSH is present there. This latter fraction
contributes insignificantly to mitochondrial AsV reduction in
normal mitochondria, but it becomes relatively more significant
and apparent upon depletion of endogenous GSH from these
organelles. Finally, the observations presented in Figure 10 are
also compatible with the proposed role of ATP synthase and
GSH in mitochondrial reduction of AsV. This figure demonstrates that GSH-depleted isolated mitochondria, when supplied with ADP and AsV to permit ADP-AsV formation
transiently before solubilizing them and supplying them with
GSH, reduced more AsV upon GSH addition than those
mitochondria in which ADP-AsV formation was prevented by
the ATP synthase inhibitors OM or AU.
Although our study strongly supports the role of ATP
synthase in mitochondrial reduction of AsV, phosphorolyticarsenolytic enzymes residing in these organelles might also be
involved. For example, it has been shown that the mitochondrial OCT, when catalyzing the arsenolysis of citrulline with
formation of carbamoyl-AsV, promotes GSH-dependent AsV
reduction (Németi and Gregus, 2009a). Purine nucleoside
phosphorylase, a predominantly cytosolic enzyme, is also
localized in mitochondria where it can promote thiol-dependent
AsV reduction when experimentally supplied with its purine
nucleoside substrate for the arsenolytic reaction that produces
ribose-1-AsV (Németi and Gregus, 2009b). However, the role
of these enzymes in mitochondrial AsV reduction is likely
insignificant normally because under physiological conditions
OCT synthesizes rather than cleaves citrulline and because the
mitochondrial purine nucleoside pool is small. Interestingly,
PNPase activity has been found in rat liver mitochondria (See
and Fitt, 1972a, 1972b) and immunostaining indentified
PNPase in the mitochondria of human cells (Piwowarski
et al., 2003; Sarkar and Fisher, 2006). This enzyme, however,
has recently been localized to the intermembrane space of
mammalian mitochondria (Chen et al., 2006, 2007; Rainey
et al., 2006). PNPase residing in the intermembrane space
could hardly contribute to AsV reduction by isolated rat liver
mitochondria via providing arsenylated nucleotides for reduction by GSH because the intermembrane space lacks RNA
(to feed the arsenolytic reaction) and because the concentration
of magnesium (which is essential for PNPase for activity) and
GSH should be very low outside the matrix space of isolated
mitochondria employed in this study. Nevertheless, further
studies are warranted to explore the conditions under which
mitochondrial PNPase might be involved in reduction of AsV
to AsIII.
Assessment of the in vivo significance of mitochondrial AsV
reduction is inherently difficult, if not impossible, especially if,
as our studies indicate, it is coupled to oxidative phosphorylation, the major process producing cellular energy. Theoretically, the high AsV reducing activity of mitochondria (Németi
and Gregus, 2002; Fig. 8) favors the idea that these organelles
contribute to reduction of AsV in tissues, whereas the competition of Pi with AsV for cell membrane and mitochondrial
transporters and reducing enzymes limits their potential role.
Recent studies have revealed that an ATP synthase is also
localized on the extracellular surface of hepatocytes and other
cells (Devenish et al., 2008; Mangiullo et al., 2008). It may be
speculated that if this ecto-ATP synthase could produce ADPAsV, it could support thiol-mediated AsV reduction unlimited
by membrane transport of AsV.
In summary, this paper presents strong, yet circumstantial,
evidence that mitochondrial ATP synthase can promote reduction of AsV. In addition, this work has added a new
member, the PNPase, to the list of five phosphorolyticarsenolytic enzymes whose AsV-reducing activity has been
demonstrated. It is proposed that both ATP synthase and
PNPase, although by different mechanisms, convert AsV into
an arsenylated nucleotide, ADP-AsV and AMP-AsV, respectively, in which the pentavalent arsenic readily undergoes
reduction by GSH to form the much more toxic AsIII.
The authors thank István Schweibert for his excellent
assistance in the experimental work.
Bertagnolli, B. L., and Hanson, J. B. (1973). Functioning of the adenine
nucleotide transporter in the arsenate uncoupling of corn mitochondria. Plant
Physiol. 52, 431–435.
Bon, S., Godefroy, T., and Grunberg-Manago, M. (1970). Cinétique des réactions
catalysées par la polynucléotide phosphorylase de Escherichia coli. Le 2désoxy-ADP comme substrat et inhibiteur. Eur. J. Biochem. 16, 363–372.
Brinkel, J., Khan, M. H., and Kraemer, A. (2009). A systematic review of
arsenic exposure and its social and mental health effects with special
reference to Bangladesh. Int. J. Environ. Res. Public Health 6, 1609–1619.
Chan, T. L., Thomas, B. R., and Wadkins, C. L. (1969). The formation and
isolation of an arsenylated component of rat liver mitochondria. J. Biol.
Chem. 244, 2883–2890.
Chen, H. W., Koehler, C. M., and Teitell, M. A. (2007). Human polynucleotide
phosphorylase: location matters. Trends Cell. Biol. 17, 600–608.
Chen, H. W., Rainey, R. N., Balatoni, C. E., Dawson, D. W., Troke, J. J.,
Wasiak, S., Hong, J. S., Mc Bride, H. M., Koehler, C. M., Teitell, M. A.,
et al. (2006). Mammalian polynucleotide phosphorylase is an intermembrane
space RNase that maintains mitochondrial homeostasis. Mol. Cell. Biol. 26,
Chou, J. Y., Singer, M. F., and Mc Phie, P. (1975). Kinetic studies on the
phosphorolysis of polynucleotides by polynucleotide phosphorylase. J. Biol.
Chem. 250, 508–514.
Csanaky, I., and Gregus, Z. (2005). Role of glutathione in reduction of arsenate
and of gamma-glutamyltranspeptidase in disposition of arsenite in rats.
Toxicology 207, 91–104.
Csanaky, I., Németi, B., and Gregus, Z. (2003). Dose-dependent biotransformation of arsenite in rats—not S-adenosylmethionine depletion impairs
arsenic methylation at high dose. Toxicology 183, 77–91.
Devenish, R. J., Prescott, M., and Rodgers, A. J. (2008). The structure and
function of mitochondrial F1FO-ATP synthases. Int. Rev. Cell. Mol. Biol.
267, 1–58.
Dilda, P. J., and Hogg, P. J. (2007). Arsenical-based cancer drugs. Cancer
Treat. Rev. 33, 542–564.
Duée, E. D., and Vignais, P. V. (1969). Kinetics and specificity of the adenine
nucleotide translocation in rat liver mitochondria. J. Biol. Chem. 244,
Ghetta, A., Matus-Ortega, M., Garcı́a-Mena, J., Dehò, G., Tortora, P., and
Regonesi, M. E. (2004). Polynucleotide phosphorylase-based photometric
assay for inorganic phosphate. Anal. Biochem. 327, 209–214.
Godefroy, T., Cohn, M., and Grunberg-Manago, M. (1970). Kinetics of
polymerization and phosphorolysis reactions of E. coli polynucleotide
phosphorylase. Role of oligonucleotides in polymerization. Eur. J. Biochem.
12, 236–249.
Gornall, A. G., Bardawill, C. J., and David, M. M. (1949). Determination of
serum proteins by means of the biuret reaction. J. Biol. Chem. 177, 751–766.
Gregus, Z., Fekete, T., Halászi, É., Gyurasics, Á., and Klaassen, C. D. (1998).
Effect of fibrates on glycine conjugation of benzoic acid. Drug Metab.
Dispos. 26, 1082–1088.
Gregus, Z., Gyurasics, Á., and Csanaky, I. (2000). Biliary and urinary excretion
of inorganic arsenic: monomethylarsonous acid as a major biliary metabolite
in rats. Toxicol. Sci. 56, 18–25.
Hungarian National Research Fund (OTKA K73228) and
the Hungarian Ministry of Health. (ETT 070/2009).
Gregus, Z., and Németi, B. (2002). Purine nucleoside phosphorylase as
a cytosolic arsenate reductase. Toxicol. Sci. 70, 13–19.
Gregus, Z., and Németi, B. (2005). The glycolytic enzyme glyceraldehyde-3phosphate dehydrogenase works as an arsenate reductase in human red blood
cells and rat liver cytosol. Toxicol. Sci. 85, 859–869.
Németi, B., and Gregus, Z. (2004). Glutathione-dependent reduction of
arsenate in human erythrocytes—a process independent of purine nucleoside
phosphorylase. Toxicol. Sci. 82, 419–428.
Gregus, Z., and Németi, B. (2007). Glutathione-dependent reduction of
arsenate by glycogen phosphorylase—responsiveness to endogenous and
xenobiotic inhibitors. Toxicol. Sci. 100, 44–53.
Németi, B., and Gregus, Z. (2007). Glutathione-dependent reduction of
arsenate by glycogen phosphorylase—a reaction coupled to glycogenolysis.
Toxicol. Sci. 100, 36–43.
Gregus, Z., Roos, G., Geerlings, P., and Németi, B. (2009). Mechanism of
thiol-supported arsenate reduction mediated by phosphorolytic-arsenolytic
enzymes. II. Enzymatic formation of arsenylated products susceptible for
reduction to arsenite by thiols. Toxicol. Sci. 110, 282–292.
Németi, B., and Gregus, Z. (2009a). Glutathione-supported arsenate reduction
coupled to arsenolysis catalyzed by ornithine carbamoyl transferase. Toxicol.
Appl. Pharmacol. 239, 154–161.
Gresser, M. J. (1981). ADP-arsenate. Formation by submitochondrial particles
under phosphorylating conditions. J. Biol. Chem. 256, 5981–5983.
Han, D., Canali, R., Rettori, D., and Kaplowitz, N. (2003). Effect of glutathione
depletion on sites and topology of superoxide and hydrogen peroxide
production in mitochondria. Mol. Pharmacol. 64, 1136–1144.
Hughes, M. F. (2006). Biomarkers of exposure: a case study with inorganic
arsenic. Environ. Health Perspect. 114, 1790–1796.
Hughes, M. F., Kenyon, E. M., and Kitchin, K. T. (2007). Research approaches
to address uncertainties in the risk assessment of arsenic in drinking water.
Toxicol. Appl. Pharmacol. 222, 399–404.
Jocelyn, P. C., and Cronshaw, A. (1985). Properties of mitochondria treated
with 1-chloro-2,4-dinitrobenzene. Biochem. Pharmacol. 34, 1588–1590.
Jones, D. H., and Boyer, P. D. (1969). The apparent absolute requirement of
adenosine diphosphate for the inorganic phosphate-water exchange of
oxidative phosphorylation. J. Biol. Chem. 244, 5767–5772.
Kitchin, K. T., and Wallace, K. (2008). The role of protein binding of trivalent
arsenicals in arsenic carcinogenesis and toxicity. J. Inorg. Biochem. 102,
Klingenberg, M. (2008). The ADP and ATP transport in mitochondria and its
carrier. Biochim. Biophys. Acta. 1778, 1978–2021.
Littauer, U. Z., and Soreq, H. (1982). Polynucleotide phosphorylase. Enzymes
15, 517–553.
Mangiullo, R., Gnoni, A., Leone, A., Gnoni, G. V., Papa, S., and Zanotti, F.
(2008). Structural and functional characterization of FOF1-ATP synthase on
the extracellular surface of rat hepatocytes. Biochim. Biophys. Acta. 1777,
Németi, B., and Gregus, Z. (2009b). Mechanism of thiol-supported arsenate
reduction mediated by phosphorolytic-arsenolytic enzymes. I. The role of
arsenolysis. Toxicol. Sci. 110, 270–281.
Páez-Espino, D., Tamames, J., de Lorenzo, V., and Cánovas, D. (2009).
Microbial responses to environmental arsenic. Biometals 22, 117–130.
Piwowarski, J., Grzechnik, P., Dziembowski, A., Dmochowska, A.,
Minczuk, M., and Stepien, P. P. (2003). Human polynucleotide phosphorylase, hPNPase, is localized in mitochondria. J. Mol. Biol. 329,
Platanias, L. C. (2009). Biological responses to arsenic compounds. J. Biol.
Chem. 284, 18583–18587.
Radabaugh, T. R., Sampayo-Reyes, A., Zakharyan, R. A., and Aposhian, H. V.
(2002). Arsenate reductase II. Purine nucleoside phosphorylase in the
presence of dihydrolipoic acid is a route for reduction of arsenate to arsenite
in mammalian systems. Chem. Res. Toxicol. 15, 692–698.
Rainey, R. N., Glavin, J. D., Chen, H. W., French, S. W., Teitell, M. A., and
Koehler, C. M. (2006). A new function in translocation for the mitochondrial
i-AAA protease Yme1: import of polynucleotide phosphorylase into the
intermembrane space. Mol. Cell. Biol. 26, 8488–8497.
Rosen, B. P. (2002). Biochemistry of arsenic detoxification. FEBS Lett. 529,
Sarkar, D., and Fisher, P. B. (2006). Polynucleotide phosphorylase: an
evolutionary conserved gene with an expanding repertoire of functions.
Pharmacol. Ther. 112, 243–263.
See, Y. P., and Fitt, P. S. (1972a). Partial purification and properties of rat liver
mitochondrial polynucleotide phosphorylase. Biochem. J. 130, 343–353.
See, Y. P., and Fitt, P. S. (1972b). A study of the localization of polynucleotide
phosphorylase within rat liver cells and of its distribution among rat tissues
and diverse animal species. Biochem. J. 130, 355–362.
Matus-Ortega, M. E., Regonesi, M. E., Piña-Escobedo, A., Tortora, P.,
Dehò, G., and Garcı́a-Mena, J. (2007). The KH and S1 domains of
Escherichia coli polynucleotide phosphorylase are necessary for autoregulation and growth at low temperature. Biochim. Biophys. Acta. 1769,
Stýblo, M., Drobná, Z., Jaspers, I., Lin, S., and Thomas, D. J. (2002). The role
of biomethylation in toxicity and carcinogenicity of arsenic: a research
update. Environ. Health Perspect. 110, 767–771.
Messens, J., and Silver, S. (2006). Arsenate reduction: thiol cascade chemistry
with convergent evolution. J. Mol. Biol. 362, 1–17.
Thomas, D. J. (2007). Molecular processes in cellular arsenic metabolism.
Toxicol. Appl. Pharmacol. 222, 365–373.
Moore, S. A., Moennich, D. M., and Gresser, M. J. (1983). Synthesis and
hydrolysis of ADP-arsenate by beef heart submitochondrial particles. J. Biol.
Chem. 258, 6266–6271.
Tietze, F. (1969). Enzymic method for quantitative determination of nanogram
amounts of total and oxidized glutathione. Anal. Biochem. 27, 502–522.
Németi, B., Csanaky, I., and Gregus, Z. (2003). Arsenate reduction in human
erythrocytes and rats—testing the role of purine nucleoside phosphorylase.
Toxicol. Sci. 74, 22–31.
Németi, B., and Gregus, Z. (2002). Mitochondria work as reactors in reducing
arsenate to arsenite. Toxicol. Appl. Pharmacol. 182, 208–218.
Tseng, C. H. (2007). Metabolism of inorganic arsenic and non-cancerous health
hazards associated with chronic exposure in humans. J. Environ. Biol. 28,
Villa-Bellosta, R., and Sorribas, V. (2008). Role of rat sodium/phosphate
cotransporters in the cell membrane transport of arsenate. Toxicol. Appl.
Pharmacol. 232, 125–134.